Analisis model partial least square dan principal component analysis dengan sektor-sektor mobilitas di beberapa negara yang memiliki hubungan asosiatif dengan covid-19

Alexandro, Juan (2023) Analisis model partial least square dan principal component analysis dengan sektor-sektor mobilitas di beberapa negara yang memiliki hubungan asosiatif dengan covid-19. Bachelor thesis, Universitas Pelita Harapan.

[img]
Preview
Text (Title)
Title.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (212kB) | Preview
[img]
Preview
Text (Abstract)
Abstract.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (573kB) | Preview
[img]
Preview
Text (ToC)
ToC.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (1MB) | Preview
[img]
Preview
Text (Chapter1)
Chapter1.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (642kB) | Preview
[img] Text (Chapter2)
Chapter2.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (792kB)
[img] Text (Chapter3)
Chapter3.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (715kB)
[img] Text (Chapter4)
Chapter4.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (2MB)
[img]
Preview
Text (Chapter5)
Chapter5.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (649kB) | Preview
[img]
Preview
Text (Bibliography)
Bibliography.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (652kB) | Preview
[img] Text (Appendices)
Appendices.pdf
Restricted to Repository staff only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (7MB)

Abstract

Masalah multikolinieritas terjadi karena satu atau lebih variabel bebas berkorelasi tinggi dengan variabel bebas lainnya. Pada umumnya, variabel tersebut memiliki pengaruh terhadap variabel tak bebas. Salah satu contoh variabel-variabel yang rentan terhadap masalah multikolinieritas adalah variabel mobilitas COVID-19. Variabel mobilitas COVID-19 dan kasus harian memiliki jumlah yang berbeda di setiap negara. Dasar pengukuran setiap variabel mobilitas harian adalah nilai median untuk hari yang sesuai selama periode 5 minggu, yaitu dari tanggal 3 Januari sampai dengan tanggal 6 Februari 2020. Data yang digunakan merupakan data yang diperoleh dari tanggal 15 Februari 2020 sampai dengan 10 September 2022. Penelitian diawali dengan membuat model regresi, lalu mendeteksi multikolinieritas pada variabel mobilitas, kemudian membentuk model regresi PCA dan PLS, dan diakhiri dengan uji asumsi klasik model regresi PCA dan PLS. Kedua model PCA dan PLS sukses menghilangkan multikolinearitas, namun meskipun model PLS sukses dalam menyelesaikan masalah uji asumsi klasik lainnya, model PLS memiliki nilai RMSE yang lebih besar daripada model PCA. Oleh karena itu, dalam penelitian ini, dapat disimpulkan bahwa model regresi PCA lebih baik mengatasi masalah multikolinieritas daripada model regresi PLS, meskipun ada sebagian negara dan periode yang memiliki nilai R2 yang kecil. / Multicollinearity problems occur because one or more independent variables are highly correlated with other independent variables. In general, these variables have an influence on the independent v ariables. One example of variables that are prone to multicollinearity problems is the COVID-19 mobility variable. COVID-19 mobility variables and daily cases have different results in each country. The basis for measuring each daily mobility variable is the median value for the corresponding day over a 5-week period, from January 3 to February 6, 2020. The data used was obtained from February 15, 2020 to September 10, 2022. The research begins by creating a regression model, then detecting multicollinearity in the mobility variable, forming PCA and PLS regression models, and ending with testing the classical assumptions of PCA and PLS regression models. Both PCA and PLS models successfully eliminated multicollinearity, however although the PLS model was successful in solving other classical assumption test problems, the PLS model had larger RMSE values than the PCA model. Therefore, in this study, it can be concluded that the PCA regression model better solves the multicollinearity problem than the PLS regression model, although there are some countries and periods that have small R2 values.

Item Type: Thesis (Bachelor)
Creators:
CreatorsNIMEmail
Alexandro, JuanNIM01112180033alexandro.juan1205@gmail.com
Contributors:
ContributionContributorsNIDN/NIDKEmail
Thesis advisorMargaretha, HelenaNIDN0312057504UNSPECIFIED
Thesis advisorCahyadi, LinaNIDN0328077701UNSPECIFIED
Uncontrolled Keywords: covid-19; multikolinearitas; principal component analysis; partial least square; mobilitas
Subjects: Q Science > QA Mathematics
Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Q Science > QA Mathematics > QA76.75-76.765 Computer software
Divisions: University Subject > Current > Faculty/School - UPH Karawaci > Faculty of Science and Technology > Mathematics
Current > Faculty/School - UPH Karawaci > Faculty of Science and Technology > Mathematics
Depositing User: Juan Alexandro Siagian
Date Deposited: 23 Aug 2023 02:14
Last Modified: 23 Aug 2023 02:14
URI: http://repository.uph.edu/id/eprint/57924

Actions (login required)

View Item View Item